next up previous
Next: About this document ...

Verena Kuhlemann
Improving the communication pattern in mat-vec operations for large scale-free graphs by disaggregation

Mathematics & Computer Science
Emory University
Suite W401
400 Dowman Drive
Atlanta
GA 30322
vkuhlem@emory.edu
Panayot S. Vassilevski

Matrix vector multiplication (mat-vec) is the key operation in any Krylov-subspace iteration method. We are interested in Krylov methods applied to problems associated with graph Laplacians arising from large scale-free graphs. Computations with graphs of this type on parallel distributed-memory computers are challenging. This is due to the fact that scale-free graphs have a degree distribution that follows a power-law and currently available graph partitioners are not efficient for such an irregular degree distribution. The lack of a good partitioning leads to excessive inter-processor communication requirements during every mat-vec. We present an approach to alleviate this problem based on embedding the original irregular graph into a more regular one by disaggregating (splitting up) vertices in the original graph. The mat-vec operations for the original graph are performed via a triple matrix product involving the embedding graph. Even though the latter graph is larger, we are able to decrease the communication requirements considerably and improve the performance of mat-vec.

This work was performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract DE-AC52-07NA27344.




next up previous
Next: About this document ...
root 2012-02-20